An Integrated Driver With Real-Time Predictive Dead-Time Optimization Technique for GaN-Based Synchronous Buck Converter

With the superior figure of merit (FOM), gallium nitride high electron mobility transistors (GaN HEMTs) currently play a critical role in the switched-mode power supplies (SMPS) application with megahertz switching frequency, high power density and efficiency. However, the frequency-dependent dead-time losses during the reverse conduction significantly degrade the conversion efficiency. Reverse conduction can also increase the risk of the bootstrap voltage overcharge. This paper proposes a dead-time optimization technique (DOT) based on predictive VSW detection that can effectively eliminate the reverse conduction. Unlike conventional methods that require multiple cycles to determine the optimal dead-time, the proposed DOT can achieve the optimal dead-time in real-time across a wide range of V<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">IN</sub> (24-48 V) and load current (0.1-6 A). A test chip was fabricated in a 0.18 μm BCD process. The experimental results demonstrate that the reverse conduction has been eliminated through zero voltage switching (ZVS) or partial ZVS with a negligible switching node voltage. Furthermore, a 4.3% efficiency improvement in 48 V to 5 V buck converter (<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">I</i><sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">L</sub> = 6 A) is achieved with the proposed DOT.